Demo: Gemma Scope: Sparse autoencoders on Gemma 2

308
11.4
Следующее
Популярные
149 дней – 4 4340:28
Night owl or a rise-n-coder? 🌙☀️
Опубликовано 18 октября 2024, 16:00
Sparse Autoencoders (SAEs) are an interpretability tool that allow you to inspect what concepts your LLM is thinking about. Learn more our open-source suite of SAEs on all layers of Gemma 2 2B and 9B, which enables everyone to analyze and steer Gemma's internals.

Watch more videos from Gemma Developer Day Tokyo → goo.gle/gemma-dev-tokyo

Subscribe to Google for Developers → goo.gle/developers

#Gemma #GemmaDeveloperDay
Свежие видео
1 день – 986 73622:11
The Nuclear Bomb Proof PC
4 дня – 3 579 64811:18
The Tri Folding Phone Impressions!
автотехномузыкадетское